It is believed that in order to fabricate integrated circuits (IC) having feature sizes below 10 nm, a process other than the lithographic processes in current use for larger feature sizes will be required. This is due in part to wavelength limitations for resolving features of that scale. Molecular electronics shows promise as the technology capable of achieving IC feature sizes of 10 nm and below. One approach to fabricating molecular electronic devices is the use of carbon nanotubes (CNT).
Carbon nanotubes have a unique property wherein they can perform as a metal or as a semiconductor, depending on configuration. Small-scale integrated circuits can take advantage of carbon nanotube sub-10 nm size and the ability to take on p- or n-type semiconductor properties. Carbon nanotubes have unique properties compared with planar semiconductor devices, including: high chemical stability; high thermal conductivity; high mechanical strength; sizes below 10 nm; semiconductor- and metallic-like properties; the prospect to regulate band-gap by changing the diameter of the carbon nanotube; the prospect to make heterojunction devices; and the prospect of vertical integration providing high density IC's.
Carbon nanotubes differ substantially in operation from planar semiconductor devices. The carbon nanotube conducts essentially on its surface where all the chemical bonds are saturated and stable. Therefore, there is no need for careful passivation of the interface between the carbon nanotube channel and the gate dielectric. In other words, carbon nanotubes have no equivalent to the silicon/silicon dioxide interface of commonly used semiconductor devices.
One major impetus to achieving success with carbon nanotube technology is the difficulty in electrically interconnecting carbon nanotubes to fabricate integrated circuits. Single CMOS transistors have been demonstrated with carbon nanotubes placed to bridge the gap between two gold electrodes which were defined lithographically on 140 nm thick SiO2 film grown on a silicon wafer. However, this method utilizing single placement of a carbon nanotube will not prove commercially viable.
Another demonstrated method involved the fabrication of gold contacts interconnecting with an array of carbon nanotubes which were grown through templates of anodized aluminum with Co or Ni catalysts placed at the bottom of the pores of anodic aluminum oxide. However, this method cannot be used to make contact between single carbon nanotubes and therefore, the carbon nanotubes can not be integrated into integrated circuits.
In order for carbon nanotube technology to be a viable approach to fabricating nanometer-scale integrated circuit devices, for use in commercial products, methods for fabricating carbon nanotube integrated circuits scalable to commercial production must be developed.